21, rue d’Artois, F-75008 PARIS
http : //www.cigre.org
114
LUND 2015
PJM Planning Study Practices and Market Impacts on the Integration of Renewable
Resources with AC-HVDC Systems
J. TONG, M. SOBOTKA, W. QIU, B. CHOI, A. KEECH
PJM Interconnection
2750 Monroe Boulevard, Audubon, PA 19403, USA
SUMMARY
Renewable resources in PJM are growing extensively and are projected to serve
approximately 20% of the load in PJM by the year 2020. This growth is due to federal and
local incentives that are driving the expansion of these resources. Great care needs to be taken
when integrating these resources into the grid as their operational characteristics differ greatly
from those of conventional thermal resources. For example, wind and solar renewable
resources in PJM are dispersed and often located far away from major load centers in the PJM
footprint. These resources have uncertain and at times uncontrollable outputs, have less
inertia, which will degrade frequency control, and in general, are incapable of voltage control
without the addition of a smart inverter device. Due to their remote location, new electric
transmission infrastructure will likely be needed to deliver these resources to the PJM load
centers for the maximum economic benefit. For optimal transmission development, it is
essential to take a long term approach that incorporates the characteristics of the sources being
connected. The transmission system must have the capacity to efficiently and reliably
aggregate these diverse renewable resources, transport usable blocks of power from remote
areas to the load centers and disperse the power to load serving entities.
The first part of this paper will discuss the PJM planning study practices on the integration of
renewable resources with AC-HVDC systems. The results show the comparison between AC
and HVDC integration focusing on losses, voltage control and power system dynamics. The
second portion of the paper will briefly discuss the potential market implications of the
efficient integration of these resources into the market using and AC-HVDC system. Several
simulations will be presented based on different cases and assumptions.
This paper includes following parts: PJM system models including the modelling of HVDC,
PJM planning study procedures, the study of the renewable resource integration with HVDC,
comparison with traditional AC, and potential market implications and conclusions.
KEYWORDS
HVDC Integration, Renewable Energy Resources, Power System Voltage Control, Power System
Stability, Energy Markets.
1
I. INTRODUCTION
Renewable energy is growing quickly in the United States and accounted for 12.9 percent of
the domestically produced electricity in 2013, and 11.2 percent of total energy generation. In
the PJM territory, renewables are projected to represent approximately 20% of the total
resources by the year 2020. Renewable resources such as wind and solar resources have very
different aspects when compared to the conventional generations; wind and solar are often
more dispersed, remote from major load centers, uncertain in their output, less inertia and
lacking voltage control. Additionally, many renewable resources that are located in remote
areas lack cost–effective access to transmission and can be deterred by the lack of coordinated
planning and permitting processes and the absence of established and clear utility regulations
to enable investments in transmission to be reimbursable (i.e., cost recovery). New electric
transmission infrastructure is needed to access recently developed renewable resources. For
optimal transmission development, it is essential to take a long term approach that reflects the
characteristics of the sources being connected. The transmission system must have capacity to
aggregate these diverse renewable sources, transport usable blocks of power to remote load
centers and disperse the power to load serving entities in an efficient and reliable manner.
As reported in [1], HVDC is a viable option to integrate renewable resources like wind power
and has several competitive advantages when compared against HVAC. The benefits of
HVDC are increased energy efficiency to help operators meet growing electricity demand
while reducing transmission losses and the amount of land required for transmission lines.
HVDC can also connect AC grids to renewable resources while improving power quality,
stability and reliability on those networks by reducing disturbances. The networks can be
managed by control centers while operators keep the grid balanced by injecting the necessary
power when there is a dip or peak in demand. One advantage that will be increasingly
valuable in the near future is the controllable nature of HVDC lines. Balancing generation and
load has become more complicated with the increased penetration of intermittent renewable
resources. [2] has reported that E.ON Netz, a high voltage grid operator, has operated HVDC
lines to provide increased flexibility to balancing generation and load. This control will allow
for increased flexibility of operating the high voltage electrical grid and will also allow for
increased penetrations of renewable resources.
The difference in cost of equipment and operation of HVAC and HVDC can be significant.
As documented in [3], HVDC lines are able to be built using a smaller right of way, to
provide more power per conductor, to occur less power losses, and to operate without reactive
compensation. However, the cost of the purchase and maintenance of the converters and AC
side filters are expensive. Depending on specific circumstances, the breakeven distance of
overhead lines between AC and DC line ranges from 500 km (310 miles) to 800 km (497
miles). Therefore, HVDC would be the most cost effective choice to transport bulk power
from remote renewable resources when the distance is approximately 650 km (400 miles) or
more.
2
II. PJM SYSTEM MODELING
a. PJM Planning System
The PJM Planning System used in the study case is the whole US Eastern Interconnection
model, including 68,607 buses and 9,470 generators. The dynamic model used for this
study includes details of generators’ excitation systems, governors, power system
stabilizers (if one exists), loads and various other equipment. The minimum modeling for
machines is a generator model. PJM currently represents loads as 100% constant current
real power and 100% constant impedance reactive power.
Thus, power system models in the study can be mathematically represented as shown
below as a set of differential equations to represent generators and the associated control
systems.
 = f ( x, y )
x

0 = g ( x, y )
(1)
( 2)
For stability analysis three phase faults, single line to ground faults with stuck breaker and
single line to ground faults with the communications failure cleared within zone 2 times
will be examined. Rotor angle, real power output, EFD, speed and terminal voltage of
units under study are monitored. Bus voltages in the same area are also monitored.
Following the disturbance, the voltage profile and oscillation envelope should meet the
requirements +/-5% of the pre-contingency voltages and 3% damping margin.
b. Conceptual Interregional HVDC Network Model
The HVDC system consists of a 3,000 MW bipolar link operating at 600 kV DC. Each
pole of the DC line connects to the PJM system through 12-pulse bridge line commutated
converters and two 3-phase transformers.
Shunt compensation is provided at both ends of the line to offset reactive power demands
of the converters determined to be about 60% of the total active power injection (1,700
MVA). Additional shunt capacitors as well as dynamic compensation were added to
properly regulate the voltage at both ends to desired values based on PJM Operating
Procedures [4].
An equivalent representation of the interconnection through a DC link is shown in Figure
1. Note that the converter transformers which facilitate coupling between the AC and DC
networks are implicitly modeled as part of the converter.
3
Figure 1. Conceptual representation of the proposed HVDC connection
The dynamic model for the DC line and the Dynamic Compensation device are based on
standard PSS/E® models populated with typical values.
III. Study of Renewable Resources with HVDC Integration
1. Static State
For the purpose of comparison in the study, a proposed set of wind generation plants are
connected to PJM transmission system through either two parallel, 500 miles long, 600
kV HVDC lines, or, two parallel, 500 mile long, 765KV HVAC lines. The AC line
parameters are shown in Table 1.
Table 1. AC line parameters
Phase Bundle
conductor Diameter (in)
Conductor Weight (lb/ft)
No. of Phase Bundles
Bundle Diameter (in)
Line Resistance %)
Line Inductive Impedance (%)
Line Charging (%)
Surge Impedance Loading, SIL (MW)
Practical loading Limit (0.68*SIL)
(MW)
Theoretical Loading Limit (MW)
6-795 kCM
ACSR
1.063
0.896
3
30
0.18
4.35
2500
2390
1625.2
3080
The voltage analysis indicates that for a 3,000MW power transfer, approximately 3,500
MVAR shunt capacitors are needed for HVDC to maintain the system voltage within the
limits; while with AC line, approximately 2,000 MVAR reactors are needed to maintain
4
the voltage since the long AC lines have significant amount of charging. For a 1,000MW
power transfer, approximately 850 MVAR shunt capacitors are needed for HVDC to
maintain the voltage; while with AC line, approximately 5,200 MVAR reactors are
needed.
At both transfer levels (3,000MW and 1,000MW), the loss of HVAC lines is a little lower
than the HVDC lines.
Table 2. Power Loss
Power Loss
600KV
HVDC
4.7%
1.5%
3,000MW power delivery
1,000MW power delivery
765KV
HVAC
3.1%
1.1%
N-1 and N-1-1 voltage studies (Voltage Magnitude and Voltage drop) were performed for
the contingencies within 5 electrical layers. The results show no significant difference
between HVDC and HVAC lines.
2. Stability
A stability study was then conducted to evaluate the dynamic performance of the HVAC
and the HVDC options to transfer 3,000MW of wind generation to PJM system. To make
the topology of the HVAC option similar to the HVDC option, a phase angle regulator
(PAR) is used as a link to the external system. It is noted that PAR is used to control the
power flow of the link in steady state condition similar to the HVDC model.
Wind
Farm
345KV
Bus
765KV
Bus
REACTOR
AC lines to
PJM system
345KV Bus
External
System
link to external system
via PAR
Dynamic
Compensation
Device
Figure 2. HVAC option with AC link to an external system via phase angle regulator
(PAR)
To evaluate dynamic performance of the two options, several contingencies and system
conditions are chosen. Baseline stability and N-1-1 stability studies are conducted using
the selected contingencies. NERC category B and C contingencies are applied to baseline
condition and NERC category B contingencies are applied to N-1 outage condition.
5
Baseline condition
For baseline condition, transient stability simulation was conducted for NERC category B
and C contingencies and maximum angle swings of generators were monitored. Table 3
shows maximum angle swings for top 5 critical contingencies.
Table 3. Maximum angle swings for several contingencies
Contingency
No.
Contingency Type
SLG* fault w/ delayed
clearing
SLG fault w/ delayed
Ctg. 2
clearing
3-phase fault w/ normal
Ctg. 3
clearing
3-phase fault w/ normal
Ctg. 4
clearing
SLG fault w/ delayed
Ctg. 5
clearing
*SLG: single-line-to-ground
Ctg. 1
Maximum angle swing (degree)
HVDC option
AC option
93
92
83
80
41
37
37
35
27
28
For the most of the tested contingencies, the HVDC option shows a slightly larger angle
swing which indicates that the HVDC option shows a comparable performance to the
HVAC option from a transient (angle) stability perspective. Sustained angle oscillation
was not observed from the tested contingencies.
N-1 outage condition
In addition to the baseline stability, N-1-1 stability was also conducted for selected N-1
(maintenance) outage condition. Under N-1 outage condition, a 3-phase fault w/ normal
contingency was applied. Table 4 shows maximum angle swings for two 3-phase faults w/
normal contingencies for an N-1 outage condition. It is observed that the tested Category
B contingencies are quite stable under the N-1 outage condition and the HVDC option still
shows a comparable performance.
Table 4. Maximum angle swings for two NERC category B type contingencies for an N-1
outage condition
Contingency
No.
Ctg. 6
Ctg. 4
Contingency Type
3-phase fault w/ normal
clearing
3-phase fault w/ normal
clearing
Maximum angle swing (degree)
HVDC option
AC option
42
17
36
35
Transient voltage recovery performance is another important aspect in stability analysis.
Figure 3 shows transient voltage recovery behaviors of the two options at 345kV bus on
6
the PJM side. To compare the voltage recovery performance, a single-line-to-ground
(SLG) fault with delayed clearing contingency due to stuck breaker was tested.
1.4
HVAC option
HVDC option
1.2
Voltage (pu)
1
0.8
0.6
0.4
0.2
0
0
0.2
0.4
0.6
1.2
1
0.8
Time (second)
1.4
1.6
1.8
2
Figure 3. Voltage comparison of 345kV bus on PJM side for a delayed clearing
contingency
At the instance of the fault the voltage drops below 0.2 pu. At the normal clearing the
remote end breaker is open which boosts the voltage up to 0.5 pu. However even after the
normal clearing the SLG fault is not cleared completely due to a local stuck breaker. The
SLG fault is completely cleared at a delayed clearing time and the voltage is finally
recovered to a pre-contingency level.
It is interesting to note that the HVDC option shows a marginally higher voltage spike
than the HVAC option at the instance of the fault clearing. The peak voltage is around 1.3
pu. This can be due to the HVDC option including shunt capacitors and that the dynamic
compensation devices are not fast enough to control the fast voltage rise at the instance of
the fault clearing.
Figure 4 shows transient behaviors of power transfer to PJM system for the same
contingency. It is observed that the HVAC option gives faster recovery of power transfer
after fault clearing. However, it also shows a power dip and that the new steady state
power transfer level is not fully recovered to a pre-contingency level. From the viewpoint
of power recovery level, the HVDC option shows better performance with slightly slower
recovery speed.
7
3500
3000
2500
HVAC option
HVDC option
Power (MW)
2000
1500
1000
500
0
-500
0
0.2
0.4
0.6
0.8
1
1.2
Time (second)
1.4
1.6
1.8
2
Figure 4. Comparison of real power transfer to PJM system for a delayed clearing
contingency
IV. PJM Market Implications
PJM executed several market simulations in an effort to measuring the impact of the
integration of HVDC systems into the RTO. The simulations were conducted for annual
periods using an hourly unit and commitment software tool. The assumption for the
simulations was that renewable resources would be integrated into PJM from remote areas
outside of the RTO footprint using a HVDC system and that an AC system would not
practically be able to be built for the same capacity and distance as the HVDC system. In
other words, the AC system would not be an option to build in order to meet the same
objective of integrating large amounts of renewable resources over long distances into PJM.
Therefore, the benefits measured from the HVDC system for the simulations conducted
represent the benefits as compared to an AC system. In addition, the analysis was conducted
assuming that all the necessary transmission upgrades required to connect the HVDC system
reliably to the PJM RTO grid were included in the base simulation. The key metrics from the
analysis were the production costs, demand costs, and the emissions tons. Several simulations
were conducted using both speculative and queued HVDC interconnections for the 2019
study year. The analysis used an interconnection of 1,000 MWs, 2,800 MWs, and 3,500
MWs of HVDC capability to develop an average change in metrics for a 3,000 MWs HVDC
line interconnection. Results of this analysis are shown in the table 5 below.
Table 5. Results of Metrics Analysis
3000 MW HVDC Interconnection
2019 Average Savings ($ millions)
Production Costs
$405
8
2019 Average Reduction (tons)
Load Payments
CO2
SOx
NOx
$181
6,178,284
11,386
4,388
The average expected production costs savings for a 3,000 MW HVDC line connected from a
location externally to PJM and terminates at a location inside of PJM is about $405 million.
In addition, assuming the transmission system is enhanced to handle the additional 3,000
MW, the payments made by load over the same period would be reduced by about $181
million not including the cost of the HVDC liune itself. Another aspect that needs to be
considered is the reduction in emissions for CO2, SOx, and NOx. The above table shows
these reductions which are a direct result of the HVDC line interconnecting 3,000 MW of
renewable resources that replace conventional higher emission generation. Reduction in
emissions will provide for a cleaner environment as well as aiding in meeting state EPA
regulations.
V. Conclusions
1. The comparison results from this study show that the HVDC system is very comparable
with the HVAC system in terms of the system security and reliability in the static state
and dynamics (transient stability and system oscillations). The HVDC can regulate the
transfer MW flow (3,000 MW) through HVDC lines into PJM system; the HVAC needs
an additional phase angle regulator (PAR) to control the power flow to achieve this 3,000
MW transfer.
2. The market impacts can be quite large but one must consider the operational, reliability,
and market components when deciding to use HVDC or conventional AC systems on the
grid. From a market’s perspective, the driving force for continued integration of HVDC
systems is ultimately the economics and weighing the integration costs, including
transmission upgrades, versus the cost savings. HVDC systems are still a relatively new
technology and with continued enhancements it may become a suitable and effective
replacement to more traditional AC systems.
BIBLIOGRAPHY
[1]
[2]
[3]
[4]
[5]
Power-Electronic Systems for the Grid Integration of Renewable Energy Sources: A
Survey Juan Manuel Carrasco, Member, IEEE, Leopoldo Garcia Franquelo, Fellow,
IEEE, Jan T. Bialasiewicz, Senior Member, IEEE, Eduardo Galván, Member, IEEE,
Ramón C. Portillo Guisado, Student Member, IEEE, Ma. Ángeles Martín Prats,
Member, IEEE, José Ignacio León, Student Member, IEEE, and Narciso MorenoAlfonso, Member, IEEE
Experience From Wind Integration in Some High Penetration Areas. Lennart S¨oder,
Member, IEEE, Lutz Hofmann, Antje Orths, Member, IEEE, Hannele Holttinen, Yihhuei Wan, Member, IEEE, and Aidan Tuohy
Comparative Evaluation of HVDC and HVAC Transmission Systems. Kala Meah*,
Student Member, IEEE, and Sadrul Ula, Senior Member, IEEE
PJM Operations Manual 3.
PJM Operations Manual 14E
9